Dynamics of Fucus serratus thallus photosynthesis and community primary production during emersion across seasons: canopy dampening and biochemical acclimation

,


Introduction
In rocky intertidal communities of temperate shores, canopy-forming macroalgae are foundation species that play a pivotal role by mitigating stressful abiotic conditions (Bulleri, 2009) and constitute highly productive systems (Mann, 1973, Niell, 1977).Sheltered shores of the Northeast Atlantic coast are characterized by stands of canopy-forming fucoids that have a distinct pattern of vertical zonation.Fucus serratus is usually the lowermost zone-forming fucoid (Chapman, 1995), recorded from Northern Portugal to Northern Norway (Jueterbock et al., 2013).Because the maximum shore height is reduced in southern populations, the actual thermal environment experienced by this cold-temperate species is similar at the center and southern edge of its distribution (Pearson et al., 2009).Southern populations, however, are less resilient to abiotic stresses (Pearson et al., 2009) and have declined over the past decades with the increase in summer sea-surface temperatures (Casado-Amezua et al., 2019).Canopy loss induced a rapid shift to turf-forming communities, leading to the functional impoverishment of the coastal system (Alvarez-Losada et al., 2020).Modeling the ecological niche of F. serratus and predicted projections of its distribution in the North Atlantic according to various climate change scenarios suggest a northward retreat of its current southern limit (Jueterbock et al., 2013).This species may thus disappear from French coasts by the end of the 21 st century if it is unable to adapt to the rising temperatures (Jueterbock et al., 2014).The disappearance of this foundation species can have catastrophic consequences for its associated communities and ecosystems.For example, community metabolism was drastically reduced after the species was experimentally removed from several rocky shores of the English Channel and the North Sea (Crowe et al., 2013, Migné et al., 2015, Valdivia et al., 2012).
The metabolism of the F. serratus stand established on a rocky shore of the south of the English Channel has been investigated in situ at several temporal scales to model its dynamics in response to light and temperature variations (Bordeyne et al., 2020).The gross community primary production (GCP) reaches 1 g C m -2 h -1 during spring and summer midday emersions (Bordeyne et al., 2015).GCP is generally limited by light availability during immersion, and rates are systematically higher during emersion than during immersion (Bordeyne et al., 2017).Overall, despite the potentially stressful conditions inherent to exposure to air, such as high light and temperature, the production of the F. serratus stand remains high at the end of the emersion periods, with GCP averaging 80% of the initial level after more than 200 min of aerial exposure (Bordeyne et al., 2017).Photoinhibition has nevertheless been measured in F. serratus at low tide (Martinez et al., 2012).Furthermore, a severe decline in photosynthesis capacity, and an accumulation of reactive oxygen species (ROS, mainly hydrogen peroxide H2O2), which can cause cell damage, have been shown in low intertidal algae under simulated emersion stress (Flores-Molina et al., 2014).However, biochemical acclimation, which protects the photochemical apparatus, may allow intertidal algae to withstand emersion stress.Such acclimation includes the production and accumulation of metabolites that function as a sunscreen, protecting the thallus from harmful solar radiations, the quenching of excess excitation energy and the scavenging of ROS (Davison & Pearson, 1996).In brown algae, phenolic compounds have been suggested to act as sunscreens (e.g.Koch et al., 2016); the xanthophyll cycle (de-epoxidation of violaxanthin to zeaxanthin) is known to efficiently promote the thermal dissipation of excess excitation energy and reduce the risk of ROS generation (Harker et al., 1999); and the ascorbate-glutathione cycle may be the main pathway for scavenging H2O2 (Nan et al., 2016).In canopy-forming macroalgae, the stress of aerial exposure may also be limited through selfcovering (Monteiro et al., 2017), preventing algae from overheating, desiccation and photoinhibition (Fernandez et al., 2015).
The multidimensional structure of macroalgae stands has been demonstrated to allow sub-optimal distribution of underwater light among thalli (Binzer & Sand-Jensen, 2002), and the temporal and spatial supplementation between thalli for light absorbance has been shown to enhance photosynthetic production of shallow water communities (Sand-Jensen et al., 2007) or intertidal communities under immersion (Tait & Schiel, 2011).Canopy protection and complementarity between thalli from different layers of the F. serratus stand have also been proposed to explain how primary production can be favored during aerial exposure under high light (Bordeyne et al., 2017), but this remains to be demonstrated.During emersion periods, the horizontal layering of thalli implies a steep attenuation of light within the canopy.Likewise, F. serratus thalli exhibited effective light harvesting at low irradiance in laboratory experiments (e.g.Lichtenberg & Kuhl, 2015).However, to understand how self-covering can enhance the primary production of the whole community, the dynamics of photosynthetic capacity has to be investigated in a naturally structured stand, in combination with an integrated measurement of the metabolism of the whole community (Tait et al., 2017).
In the present survey, the aerial production of the community was assessed in situ by measuring carbon fluxes inside closed chambers.We assessed the photosynthetic activity of thalli from different layers of the canopy concurrently by measuring chlorophyll a fluorescence.In addition, we evaluated the involvement of biochemical compounds in photoprotective mechanisms, the induction of ROS and the activation of the antioxidant system by quantifying the phenolic compounds, xanthophyll cycle pigments (violaxanthin, antheraxanthin and zeaxanthin), H2O2 and ascorbate and glutathione present in thalli.The study aimed to 1) describe the in situ dynamics of the F. serratus community primary production and the photosynthetic performance of thalli throughout the midday emersion period under various environmental conditions (according to the season and the weather); 2) investigate the biochemical acclimation patterns of F. serratus across seasons and during emersion.We tested the hypothesis that, depending on the prevailing environmental conditions, thalli at the top of the canopy develop photoprotective and/or scavenging mechanisms.

Study site, environmental conditions and experiments schedule
This study was performed on the shore located in front of the Roscoff Marine Station (48°43.743'N,3°59.407'W) in the southwestern part of the English Channel.This location experiences seasonal variations in environmental parameters typical of the temperate part of the Northern Hemisphere (Bordeyne et al., 2015).Photosynthetically available radiation (400-700 nm, PAR) and air temperature are recorded continuously by a weather station (flat quantum sensor SKP215, Skye Instruments LTD) located on the building of the Roscoff Marine Station and averaged over 15 min.Seawater temperature and nutrient concentrations are measured fortnightly at the Roscoff-Estacade sampling point of the SOMLIT network (data available online at https://somlit.fr/visualisation-des-donnees/).
At the low mid-intertidal level of the shore, the semi-sheltered boulder reef is typically characterized by a dense stand of the fucoid Fucus serratus (about 70 ind m -2 accounting for 2 to 13 kgFW m -2 according to the season, Bordeyne, 2016).During emersion periods, the long F. serratus thalli (averaging 38 cm, Munda & Kremer, 1997) form a layered canopy which height can reach several decimetres.During spring tides, the middle part of this stand (about 3 m above chart datum) is exposed to air for about 4 h around noon.Two types of experiment were carried out during the emersion period of spring tides, occurring around midday under various environmental conditions (Table 1).In the first type of experiment, photosynthesis was compared between thalli at the top of the canopy and thalli at the bottom of the canopy: two experiments were performed under high and relatively constant light (Exp A the 26 th May and Exp B the 22 nd August 2017), two under high but more variable light (Exp C the 27 th May and Exp D the 23 rd August 2017) and one under low light (Exp E the 04 th December 2017).In the second type of experiment, photosynthesis was compared between thalli at the top of the canopy and thalli in intermediate layers: one experiment was performed under high and relatively constant light (Exp F the 13 th August 2018) and one under low light (Exp G the 05 th December 2017).
During the emersion period of each date of experiments, PAR was measured at the level of the substratum using a flat quantum sensor (LiCor SA-190) and recorded every minute.PAR reaching thalli at the bottom of or within the F. serratus canopy was measured at the same time as fluorescence using a mini flat quantum sensor (Walz).The air temperature was measured at the Fucus stand and recorded every minute using two HOBO loggers placed respectively above (Tab) and below (Tbe) the canopy.A canopy temperature buffering index (TBI) was then calculated every minute as TBI = (Tab -Tbe) / Tab.

Community primary production
Productivity was assessed at the F. serratus community scale by measuring carbon dioxide (CO2) fluxes inside benthic chambers.At the beginning of emersion period, three chambers were haphazardly positioned (approximately 2 m apart) to account for spatial variability.Each chamber (made of a transparent Perspex® dome with a 0.3 x 0.3 m transparent air-tight Perspex® base, total volume 17.7 L) was sealed to the substratum (using sand) and was connected to an infrared CO2 gas analyzer (LiCor Li-820) in a closed air circuit (with a flow of about 1 L min -1 ).CO2 air concentration (μmolCO2 molair -1 ) was recorded every 15 s for 5 to 10 min during incubation to calculate CO2 fluxes as described in Migné et al. (2002).Measurements were performed at ambient light and in darkness by covering the chambers to estimate net community production (NCP) and community respiration (CR), respectively.Benthic chambers were opened between two consecutive incubations to renew the ambient air.Gross community production (GCP), calculated as the sum of NCP and CR, was expressed in carbon units (mg C m -2 h -1 ) assuming a molar volume of 22.4 L mol -1 at standard temperature and pressure and a molar mass of 12 g C molCO2 -1 .Successive sets of light and dark incubations were carried out from the onset of emersion period to the return of seawater at intervals of ca. 30 min.

Thalli photosynthesis
Photosynthesis of F. serratus thalli was assessed in situ using a pulse-amplitude modulated (PAM) fluorometer (Diving PAM, Walz).The fluorescence signal was always taken from the same place in the middle of the thallus for three individuals, haphazardly selected among thalli from the top of the canopy and three other individuals from the bottom of the canopy or from an intermediate layer of the canopy.Thalli were selected according to the incident irradiance reaching them: approximately 1% and 20% of incident PAR at the bottom and within the intermediate layer of the canopy respectively.The effective quantum yield of photosystem II (ΦPSII) was measured under ambient light.The optical fibers were mounted in a home-made holder that kept the distance between the fiber optics and the algal tissue constant and standard, with a 60° angle, avoiding shading or darkening (Figure 1).ΦPSII was calculated as (Fm'-Ft)/Fm' (Genty et al., 1989), where Fm' is the maximal level of fluorescence measured during a single saturating pulse of white light (0.8 s), and Ft is the steady-state level of fluorescence measured immediately before the flash, under actinic illumination.ΦPSII estimates the relative electron transport rate (rETR in μmol electrons m -2 s -1 , hereafter referred to as μmol e -m -2 s -1 ) as ΦPSII × PAR × 0.5, where PAR is the photosynthetically available radiation (in μmol photons m −2 s −1 ), and 0.5 is a correction factor based on the assumption that the incident photons are absorbed equally by the pigments of the two photosystems.rETR was measured from the onset of emersion period to the return of seawater at intervals of ca. 30 min.The optimal quantum yield of PSII photochemistry was measured on darkadapted thalli and was calculated as Fv/Fm = (Fm -F0)/Fm (Genty et al., 1989), where Fv is the variable fluorescence, F0 is the minimal level of fluorescence, and Fm is the maximal fluorescence obtained during the application of a saturating pulse of white light (0.8 s).Thalli were dark-adapted by applying a leaf clip on the samples for 10 min, a period estimated to be long enough to allow the complete re-oxidation of the primary quinone electron acceptor of PSII (Kalaji et al., 2014).Fv/Fm assesses the extent of photoinhibition (Maxwell & Johnson, 2000) and was measured at the beginning, middle and end of the emersion period.
Relative water content of thalli Six haphazardly selected F. serratus thalli were detached from the rock at the onset of emersion to monitor their relative water content (RWC) during the emersion period.Three were placed at the bottom of or inside the canopy, and three were placed at the top of the canopy.The weight (W) of each thallus was measured in the field at the beginning, middle and end of the emersion period.Brought back to the laboratory, the thalli were rehydrated overnight to assess their fresh weight (FW) and then dried for 48 h at 60°C to assess their dry weight (DW).The relative water content of a thallus was calculated as the part of total water content [(W-DW)/(FW-DW)] at the beginning, middle and end of emersion.Thallus samples were taken at the beginning, middle and end of the emersion period, from three individuals haphazardly selected at the top and three thalli haphazardly selected from the bottom (or in an intermediate layer) of the canopy, and immediately placed in the dark and frozen in liquid nitrogen until further biochemical analysis.
For the analysis of chlorophyll and xanthophyll cycle pigments, thallus samples (disks of 8 mm diameter) were first gently patted dry to remove epiphytes.Pigments were extracted by grinding the disks in a cold mortar with methanol and small drops of methylene chloride under dim light.Extracts were centrifuged (5 min, 13 000 rpm) and supernatants were collected and filtered on polytetrafluoroethylene membranes (0.2 µm) and dry-evaporated under nitrogen.Salt contents of the extract were removed from the pigment solution in a methylene chloride:distilled water mixture (50:50, v/v) (salts stay in the aqueous phase, and pigments are found in the organic phase).The organic phase was then evaporated with nitrogen and dissolved again in 40 µL methanol just before injection.Pigment analysis was performed using high performance liquid chromatography (HPLC) (Shimadzu, Nexera XR) with a reverse-phase column (C18 Allure, Restek).Briefly, 20 µL were injected, and separation was carried out with a solvent delivery profile adapted from Arsalane et al. (1994).The conversion of violaxanthin (V), a pigment with no photoprotective properties into antheraxanthin (A) and zeaxanthin (Z), which are involved in the dissipation of energy into heat (Bilger & Bjorkman, 1990), was estimated by calculating the de-epoxidation ratio: DR = (A + Z)/(V + A + Z).
For the analysis of the other compounds, thallus samples (pieces of about 10 gFW) were patted dry, lyophilized and grounded.Phenolic compounds were determined using a protocol adapted from Ratkevicius et al. (2003) and from Contreras et al. (2005): 0.05 g of ground lyophilized thallus samples were homogenized with a glass pestle in an Eppendorf tube containing 1 mL phosphate buffer 0.1 M pH = 7.Homogenates were centrifuged (10 min, 13 000 rpm) and aliquots of 100 µL of the supernatants were added to a reaction mixture containing 3% of sodium carbonate and 0.3 M Folin-Ciocalteau reagent in a final volume of 1 mL.After an incubation period of 2 h at room temperature, the absorbance was measured at 765 nm (UV-2450, UV-VIS, Shimadzu).Total phenolic compounds (PC) were expressed in g of nano-equivalents of gallic acid per 100 g of DW (% DW) using a calibration curve prepared with gallic acid.Hydrogen peroxide (H202) concentrations were determined using a protocol adapted from Lee and Shin (2003): 0.05 g of ground lyophilized thallus samples was homogenized with a glass pestle in an Eppendorf tube containing 1 mL of trichloroacetic acid (TCA).Homogenates were centrifuged (10 min, 13 000 rpm) and aliquots of 0.2 mL of supernatant were added to a reaction mixture containing 0.2 mL of phosphate buffer and 0.8 mL of potassium iodide.After an incubation period of 1 h in darkness, absorbance was measured at 390 nm (UV-2450, UV-VIS, Shimadzu).The equivalent concentration was obtained by using a standard curve prepared with TCA and a solution of 0.01 M H2O2.Ascorbate (Asc) concentrations were determined using a protocol adapted from Hodges et al. (1996): 0.05 g of ground lyophilized thallus samples was homogenized with a glass pestle in an Eppendorf tube containing 1 mL of 2.5 M perchloric acid.Homogenates were centrifuged (10 min, 13 000 rpm) and aliquots of 0.1 mL of supernatant were added to 5 μL of 100 mM dithiothreitol.After an incubation period of 1 h at room temperature, the reaction was stopped by adding 5 μL of 5% n-ethylmaleimide and 900 μL of a reaction mixture containing 2% TCA, 8.8% orthophosphoric acid, 0.2% 2-2'-bipyridyl and 10 mM iron chloride and then incubated for 1 h at 40°C.Absorbance was measured at 525 nm (UV-2450, UV-VIS, Shimadzu).The equivalent concentration was obtained by using standard curves prepared with derivative L-AA standard.Glutathione (Glu) concentrations were determined using a protocol adapted from Hodges et al. (1996): 0.05 g of ground lyophilized thallus samples was homogenized with a glass pestle in an Eppendorf tube containing 1 mL of 5% 5-sulfosalicylic acid dehydrate.Homogenates were centrifuged (10 min, 13 000 rpm) and aliquots of 0.1 mL of supernatant were added to 150 μL of 500 mM phosphate buffer (pH 7.5).The neutralized extract was added to a reaction mixture containing 100 mM phosphate buffer (pH 7.0), 0.15 mM NADPH, 60 μM DNTB and 0.66 U glutathione reductase for the quantitative analysis of total Glu and then incubated for 1 h at 37°C.Absorbance was measured at 412 nm (UV-2450, UV-VIS, Shimadzu).The equivalent concentration was obtained by using standard curves prepared with an L-Glutathione reduced BioXtra (GSH) standard.

Data analysis
Pearson correlation was tested between the canopy temperature buffering index (TBI) and the mean temperature above canopy (Tab).A principal component analysis (PCA) was performed on the mean values of biochemical parameters (DR, H2O2, Asc, Glu), relative water content (RWC), and optimal quantum yield (Fv/Fm) measured at the end of the emersion period on the different experiment dates to explore a seasonal pattern.The seasonal acclimation of F. serratus was also investigated based on phenolic contents (PC) and xanthophyll cycle pigments (V + A + Z), normalized to the chlorophyll a content, for each of the four months of measurement (May, August and December 2017 and August 2018).Data obtained from thalli sampled at the beginning of emersion on two consecutive days were pooled.Kruskal-Wallis tests (non-parametric equivalent to ANOVA) and post-hoc pairwise Wilcoxon tests (with the Holm method for p-value adjustment) were used to analyze differences between the four months of measurement.

Environmental conditions
Across the whole experiment (May 2017-August 2018), air temperature recorded by the weather station ranged from -3.2 to 26.2 °C.Monthly mean air temperature varied seasonally (16.5 °C in August 2017, 6.3 °C in February 2018and 17.5 °C in July 2018).Samples from the Roscoff-Estacade point of the SOMLIT network showed typical seasonal variation in monthly mean seawater temperature (16.4 °C in August 2017, 9.3 °C in February 2018and 16.3 °C in August 2018) and in monthly mean inorganic nitrogen concentration (0.29 ± 0.01 μmol L -1 in May 2017, 9.40 ± 2.06 μmol L -1 in January 2018 and 0.74 ± 0.59 μmol L -1 in July 2018).
During the experiments, mean air temperature measured just above the F. serratus canopy (Tab) varied from 12.3 to 32.9 °C and the mean canopy temperature buffering index (TBI) varied from 0.06 to 0.35 (Table 1).TBI was positively correlated with Tab (r = 0.928, n = 7, p < 0.01).Incident PAR measured at the level of the substratum varied according to the season and weather conditions (Figures 2 & 3), with a mean in the range of 200-1859 μmol m -2 s -1 (Table 1).

Community primary production
Mean gross community primary production (GCP) higher than 1 g C m -2 h -1 was measured at the beginning of emersion on three dates (Figures 2 & 3): 1.28 g C m -2 h -1 on 26 May 2017 (Exp A), 1.26 g C m -2 h -1 on 22 August 2017 (Exp B) and 1.03 g C m -2 h -1 on 13 August 2018 (Exp F).On these dates, which are the ones with the highest temperature and incident light (Table 1), GCP decreased during the emersion period, representing respectively 66, 74 and 59% of the initial value at the end of the emersion period.On 27 May (Exp C), 23 August (Exp D) and 05 December 2017 (Exp G), GCP remained relatively stable during the emersion period (averaging 0.91, 0.80 and 0.43 g C m -2 h -1 , respectively).On 04 December 2017 (Exp E), GCP varied according to the incident light (reaching 0.92 g C m -2 h -1 when PAR peaked at 800 μmol m -2 s -1 ).

Thalli photosynthesis
The highest relative electron transport rate (rETR) averaged 247 μmol e -m -2 s -1 and was measured in thalli at the top of the canopy at the beginning of emersion on 26 May 2017 (Exp A, Figure 2).In thalli at the top of the canopy, rETR dramatically decreased during emersion in May and August (Exp A, B, C, D, Figure 2 & Exp F, Figure 3).Mean rETR decreased to values lower than 20 μmol e -m -2 s -1 due to very low values of effective quantum yield: ΦPSII was less than 0.05 (and even less than 0.01 on 13 August 2018, Exp F).In December, ΦPSII remained higher than 0.60, and rETR varied with the incident light on 04 December (reaching 145 μmol e -m -2 s -1 when PAR peaked at 800 μmol m -2 s -1 , Exp E, Figure 1) and remained relatively stable on 05 December (averaging 65 μmol e -m -2 s -1 , Exp G, Figure 2).In thalli at the bottom of the canopy, ΦPSII averaged 0.75 across all measurement dates, but rETR remained virtually nil (Figure 3) due to the very low PAR (Table 1).In thalli in the intermediate layers of the canopy, ΦPSII averaged 0.72 in August and 0.75 in December, and rETR averaged 53 and 14 μmol e -m -2 s -1 in August (Exp F) and December (Exp G), respectively (Figure 3).In August 2018, rETR in the intermediate layers of the canopy was higher than at the top of the canopy for most of the emersion period.The optimal quantum yield (Fv/Fm) measured in thalli at the top of the canopy decreased during emersion period in May and August (Exp A, B, C, D, Figure 4 & Exp F, Figure 5), even reaching a null value in August 2018 (Exp F), and remained stable (higher than 0.73) in December (Exp E, Figure 4 & Exp G, Figure 5).In thalli at the bottom of the canopy (Figure 4) and in the intermediate layers of the canopy (Figure 5), mean Fv/Fm always remained higher than 0.65.

Biochemical parameters
The de-epoxidation ratio (DR) tended to decrease during the emersion period on 26 May 2017, notably at the bottom of the canopy (Exp A, Figure 4).No trends were observed for DR on the other days (Exp B, C, D, E, Figure 4; Exp F, G, Figure 5).Globally, mean DR was relatively high in May and August (0.45 ± 0.02) and low in December (0.20 ± 0.02).
The principal component analysis (PCA), performed on the mean values of biochemical parameters (as well as of RWC and Fv/Fm) measured at the end of the emersion period highlighted the strong correlation between Fv/Fm and RWC (r = 0.938, n = 14, p < 0.001, Figure 6a).A seasonal pattern appeared in thalli at the top of the canopy (Figure 6b), with high DR characterizing summer samples (Exp B, D, F) and high Asc and Glu characterizing spring samples (Exp A, C).
The phenolic content (PC) of F. serratus thalli varied significantly according to the month of measurement (Kruskal Wallis test, p < 0.001), with a significantly higher mean value in December 2017 (3.31% DW, post-hoc test, p < 0.01) than in any other month of measurement (Figure 7a).The pool of pigments involved in the xanthophyll cycle normalized to the chlorophyll a content (VAZ/Chla) varied significantly with month of measurement (Kruskal Wallis test, p < 0.001), the highest value was observed in May 2017 (0.23), and the lowest in December 2017 (0.15, Figure 7b).

Community primary production and photosynthesis
The studied intertidal Fucus serratus stand exhibited high aerial production rates (sometimes exceeding 1 g C m -2 h -1 ) under high light and temperature (Figures 2 & 3).During the emersion period, production rates never decreased to less than 59% of the initial value.This pattern was already highlighted in a previous study (Bordeyne et al., 2017), but our study demonstrates that production is maintained at a relatively high rate due to the dampening effect of the canopy on temperature and light and to various acclimation mechanisms.Canopy shading helps prevent the thallus from overheating and retains humidity during emersion, and selfcovering in intertidal stand of macroalgae has been considered as an intraspecific facilitation mechanism that protects the algae from photoinhibition (Fernandez et al., 2015).The temperature buffering index (TBI) measured here was positively correlated with air temperature.Under the harshest conditions of the present survey, TBI was comparable to that measured on more southern Fucus species stands, situated at higher shore levels (Monteiro et al., 2017).We did not measure the humidity buffering index here, but regardless of the encountered weather conditions, the relative water content (RWC) was higher in thalli placed below the canopy than in thalli at the top of the canopy, confirming the canopy dampening effect.The previously proposed hypothesis of intraspecific facilitation as a mechanism regulating the production of intertidal macroalgae stands (Bordeyne et al., 2017) is validated by the physiological measurements presented here.During emersion under mild conditions, photosynthesis, measured as the electron transport rate (rETR), varied in thalli at the top of the canopy in response to changing incident light (Exp E, Figure 2).This variation confirms the capacity of the species for short-term photoacclimation, as already shown in laboratory experiments (Lichtenberg & Kuhl, 2015).In contrast, during emersion under harsh conditions, the efficiency of photosystem II photochemistry dramatically decreased in thalli at the top of the canopy.Under these conditions, the efficiency of the photosystem II photochemistry nonetheless remained high in thalli at the bottom of or within the canopy.Photosynthesis was limited in the lowest layer of thalli due to self-shading which prevented light from reaching them (only approximately 1% of the incident irradiance reached the bottom of the canopy), but was effective in the intermediate layers under relatively low local irradiance (approximately 20% of incident irradiance).In the Roscoff F. serratus stand, thalli are typically long, wide, ramified and deeply serrated (Munda & Kremer, 1997).This morphology helped distribute the available irradiance within the canopy and also protected the photosynthetic tissues within the canopy from emersion stress.The unpredictable nature of the distribution of light within the canopy nevertheless involved acclimation of the photosynthetic apparatus.The optimal photosynthetic quantum yield (Fv/Fm) also dramatically decreased during emersion periods of spring and summer in thalli at the top of the canopy, but not in thalli beneath the canopy (at the bottom, Figure 4, or in intermediate layers, Figure 5).However, photoinactivation kinetics varied with the prevailing weather conditions.The strong correlation between Fv/Fm and RWC measured in thalli at the end of the emersion period (Figure 6a) suggests that photoinactivation of PSII is mainly governed by thallus dehydration.Photoinactivation of PSII by desiccation has been proposed as a mechanism to prevent an overload of the photosynthetic apparatus in intertidal algae under high light during emersion at low tide, therefore, protecting them from photodamage (Huppertz et al., 1990).Laboratory experiments performed under high light and high temperature (Fernandez et al., 2015, Martinez et al., 2012) suggest that the recovery of the photosynthetic capacity of F. serratus thalli exposed during emersion may be not complete before the onset of the next midday low tide.The same thalli are, however, unlikely to be at the top of the canopy in consecutive low tides and inhibition can be reversed by self-shading.Fv/Fm at the top of the canopy reached a null value at the end of the emersion period on 13 August 2018 (Exp F, Figure 5).Contrarily to August 2017 (Exp B), harsh conditions, notably high air temperatures, occurred not only the day of measurement on 13 August 2018 (Exp F), but also the previous days, and we hypothesize a possible cumulative effect due to overheating during this spring-tide period.Canopy protection may therefore not be sufficient to allow this intertidal alga to withstand several consecutive harsh midday conditions during a spring-tide cycle.

Biochemical acclimation
Significant changes in phenolic content and pigment content between the different periods of measurements (Figure 7) indicate a seasonal biochemical acclimation of F. serratus.Canopy-forming algae, in particular fucoids, are known to be rich in phenolic compounds, which have been suggested to act as a sunscreen against harmful solar radiations (high PAR and UV).Therefore, we expected to observe a seasonal pattern in phenolic compounds in F. serratus thalli, with maximum values in summer as previously observed in the same geographical area (Connan et al., 2004).However, phenolic compounds have various functions (including protection against grazers and epiphytes) which may lead to different seasonal trends.Furthermore, environmental factors may affect their concentration in brown algae (Ragan & Jensen, 1978, van Hees et al., 2017).For example, both field surveys and experimental studies on intertidal seaweeds suggest that the synthesis of phenolic compounds depends on nitrogen availability (Yates & Peckol, 1993) or is inhibited by high temperatures (Mannino et al., 2016, Mancuso et al., 2019).Here, as in a Norwegian Fucus vesiculosus population (Ragan & Jensen, 1978), high contents were measured under low temperature and high nitrogen availability (in December), and low contents were measured under high temperature and low nitrogen availability (in May and August), which does not support the idea of a photoprotective role for phenolic compound accumulation.Because the xanthophyll cycle is the principal non-photochemical quenching mechanism in brown algae (Harker et al., 1999), we expected to observe an accumulation of xanthophyll pigments in spring and summer.The xanthophyll pigment content normalized to the chlorophyll a content measured in December (0.15) appeared nevertheless relatively low compared with the values (0.23-0.26) reported for a population sampled in winter in southwest England (Nielsen & Nielsen, 2010).In that more northern population, the phenolic content was, in contrast, particularly low (between 0.50 and 1.40% DW).The opposite trends in the xanthophyll pigments content and phenolic content observed here between seasons, as well as between the two populations in winter, indicate that there may be complementary photoprotection mechanisms and different strategies in different F. serratus populations.In intertidal fucoids, the photoprotective role of phenolic compounds may depend not only on their sunscreen effect, but also on their antioxidant capacities (Connan et al., 2007), and zeaxanthin may also have antioxidative properties (Nielsen & Nielsen, 2010).
Violaxanthin de-epoxidation is a rapid protection mechanism of the photosynthetic apparatus by excess light energy dissipation (Hanelt, 1996), and we expected this mechanism to be activated in thalli exposed to high light at the onset of emersion.Relatively high de-epoxidation ratio (DR) values were however measured not only in thalli at the top of the canopy, but also in thalli beneath the canopy at the beginning of the emersion period in spring and summer (Figures 4 & 5).These high values suggest that the mechanism is activated as a response to supersaturating light irradiance in very low water depth at the end of immersion.The high correlation between the DR values at the beginning of emersion and the prevailing PAR just before emersion (Pearson correlation r = 0.950, n = 7, p < 0.001, Appendix 2) tends to confirm this hypothesis.Although high temperature and dehydration may have made additive contributions to stress, DR did not increase with decreasing Fv/Fm in thalli lying at the top of the canopy during emersion.In contrast, desiccation and high temperature may have lowered DR values, as shown in the uppermost midlittoral species Pelvetia canaliculata (Harker et al., 1999, Fernandez-Marin et al., 2011).The ascorbate-glutathione cycle may then compensate for the limited capacity of the xanthophyll cycle, inducing an increase in the size of ascorbate and glutathione pool, as experimentally demonstrated in the brown alga Sargassum thunbergii (Nan et al., 2016).The simultaneous analysis of biochemical and photosynthetic parameters measured in thalli at the top of and beneath the canopy at the end of emersion (Figure 6) suggests that coordination between the xanthophyll and the ascorbateglutathione cycles varies across seasons rather than during low tide.In general, no changes in ascorbate and glutathione content were observed during the emersion period; the occasional changes occurred in thalli both at the top of and beneath the canopy, suggesting a response to environmental conditions prevailing before emersion.Accumulation of hydrogen peroxide (H2O2) was nevertheless observed on one occasion at the top of the canopy, indicating that the ascorbate and glutathione pools were insufficient to scavenge H2O2 during this particular emersion period on 13 August 2018 (Exp F).

Conclusion
Measuring the photosynthetic performance of isolated thallus pieces, Jueterbock et al. (2014) suggested that the cold-temperate F. serratus physiology lacks the plasticity to respond to the thermal extremes predicted in the near future.The present study shows the importance of the dampening canopy effect as well as biochemical acclimation for the physiology of this species and its ability to withstand environmental stresses.F. serratus canopies alter the environmental conditions and mitigate the levels of stress experienced in the lower layers of the same stand during emersion, whereas various acclimation strategies allow the adjustment of its photosynthetic capacity.Oxidative stress was nevertheless recorded once, presumably in response to recurrent heat stress.The potential of this foundation species to respond to frequent heat waves -predicted to occur in the future as a result of climate change -remains to be determined.If unable to adapt, the species could disappear, that would have catastrophic consequences on the whole community leading to the impoverishment of the coastal system.

Figure 1 .
Figure 1.In situ survey of fluorescence of thalli beneath the canopy

Figure 5 .
Figure 5. Mean (± se, n = 3) optimal quantum yield of PSII (Fv/Fm), relative water content (RWC) and de-epoxidation ratio (DR) measured during emersion on thalli at the top and intermediate layers of the Fucus serratus canopy on different dates

Figure 6 .Figure 7 .
Figure 6.Projection onto the PC1-PC2 plane (a.Variables map, b.Objects map) of the results of principal component analysis (PCA) performed on the mean values of biochemical parameters (deepoxidation ratio (DR), hydrogen peroxide (H2O2), ascorbate (Asc), and glutathione (Glu) contents), relative water content (RWC), and optimal quantum yield (Fv/Fm) measured at the end of the emersion period on May (A, C), August (B, D, F) and December (E,G) in thalli from different layers of the Fucus serratus canopy: top (white), bottom (dark) and intermediate (grey)

Appendix 2 .
Relationship between de-epoxidation ratio at the beginning of emersion (DRi, mean ± se, n = 6) and the incident light during the hour preceding emersion (as mean photosynthetically available radiation, PAR, n

Table 1 .
Environmental conditions for the two types of experiments, comparison between thalli at the top and the bottom of the canopy (I) and at the top and in intermediate layer of the canopy (II).Low tide time and height (above chart datum) and total emersion period at the 3 m shore level predicted in Roscoff on each experiment date (YY/MM/DD).Light and temperature recorded at different layers in the Fucus serratus canopy during the measurement period.UT: universal time, PAR: photosynthetically available radiations, rETR: relative electron transport rate, TBI: canopy temperature buffering index measured in thalli from different layers of the canopy (top, bottom or intermediate) on seven dates